Miryung Kim访问实验室并做学术报告

Title: Analysis and Automation of Systematic Program Modifications

Software modifications are often systematic. They consist of similar, but not identical, program changes to multiple contexts. Making such repetitive modifications is a tedious and manual process. A failure to update all relevant change locations could lead to costly errors of omissions. Existing approaches for automating and analyzing systematic changes are significantly limited. Refactoring engines are confined to a predefined set of semantics-preserving transformations, and source transformation tools require programmers to prescribe edits in a formal syntax in advance. I will present the design and implementation of a program transformation tool called SYDIT that generates a context-aware, abstract edit script from example program differences in order to automatically apply similar edits to relevant code locations.

In this talk, I will also present a field study of refactoring rationale, benefits, and challenges at Microsoft. It is widely believed that refactoring improves software quality and developer productivity. However, few empirical studies quantitatively validate refactoring benefits or investigate developers' perception towards refactoring benefits and risks. Our study found that current refactoring engines are insufficient and do not address the various technical challenges of initiating and conducting refactoring.  A designated re-architecting team in Windows overcame some of these challenges by leading centralized and top-down re-architecting effort and by developing custom tools and processes. A quantitative analysis using Window 7 version history found that there is a visible effect of refactoring---the re-architecting effort tamed the growth of binary level dependencies and contributed to reduction of post-release defects.
 
Bio: Miryung Kim is an Assistant Professor at the University of Texas at Austin. She received her M.S. and Ph.D. in Computer Science and Engineering from the University of Washington under the supervision of David Notkin in 2003 and 2008 respectively. Before that, she received an undergraduate degree in Computer Science from Korea Advanced Institute of Science and Technology (KAIST). Her current research interests lie in automated software analysis algorithms and development tools to improve developer productivity during software evolution.  In particular, her research focuses on program differencing, refactoring, and automated program transformation. She is a recipient of 2011 Microsoft Software Engineering Innovation Foundation (SEIF) Award, 2009 IBM Jazz Innovation Award, and ACM SIGSOFT Distinguished Paper Award at ASE 2010.